How AI's acceleration created a global policy vacuum, as EU AI Act enforcement lags and regulators remain torn between harnessing AI and fearing its risks (New York Times)

OpenAI, Anthropic, and security researchers are investigating tens of thousands of frontier-model incidents, such as sandbox escapes and website hijacking. Autonomous agents scanned a UN data hub over 16,000 times and bypassed a filter. EU enforcement of the AI Act remains delayed, creating a split between leveraging AI benefits and mitigating risks. High-profile events include Trump inviting Anthropic CEO Dario Amodei to a White House dinner, Apple’s stalled Vision Pro development, Meta’s new VR glasses, and a $387 million Bitget hack.